KR20020026575A - Silicon wafer and method for manufacture thereof, and method for evaluation of silicon wafer - Google Patents

Abstract

A method of manufacturing a silicon ingot by pulling up a silicon single crystal according to the Czochralski method in order to provide a method of manufacturing a silicon wafer having sufficient characteristics for a semiconductor device while doping nitrogen, And a silicon ingot is pulled up under the condition that the nitrogen concentration is 5 × 10 ¹³ atoms / cm ³ to 1 × 10 ¹⁵ atoms / cm ³ to form a silicon ingot, wherein the nitrogen concentration is 5 × 10 ¹³ atoms / cm < ³ > to 1 x 10 < ¹⁵ > atoms / cm < ³ >

Description

TECHNICAL FIELD [0001] The present invention relates to a silicon wafer, a method of manufacturing the same, a method of evaluating a silicon wafer,

As a silicon wafer (CZ-silicon wafer) manufactured by the Czochralski method (hereinafter referred to as CZ method), void defects due to vacancies detected as, for example, LSTD, FPD and COP exist . Since such a crystal defect adversely affects the quality of the final product, a heat treatment such as a hydrogen atmosphere high-temperature heat treatment (also referred to as hydrogen annealing or hydrogen annealing) may be performed to remove it. Actually, it has been known that CZ-silicon wafers subjected to hydrogen treatment disappear void defects due to vacancies detected as LSTD or the like, and exhibit excellent oxide film breakdown voltage characteristics (Japanese Patent Application Laid-Open No. Hei 3 (1991) -80338) .

However, there is a problem that the effect of the hydrogen heat treatment is limited only to the vicinity of the extreme surface of the wafer. Therefore, attention is focused on the fact that the smaller the defect size, the greater the decaying effect of defects due to the hydrogen heat treatment. There has been proposed a method in which the defect size is made finer and the hydrogen heat treatment is applied to the defect so that the effect of the heat treatment of the hydrogen is achieved until reaching the deep portion (Japanese Patent Application Laid-Open No. 10 (1998) -208987) .

In order to cope with the increase in the crystal diameter in this method, a V / G (V: pulling rate, G: crystal axis direction temperature near the melting point (Japanese Patent Application Laid-Open No. 10 (1998) -260666).

However, recently, as another approach to defect size reduction, it has been reported that the size of a defect is reduced by adding nitrogen at the time of CZ-silicon single crystal growth, and it is possible to manufacture a wafer suitable for annealing (Japanese Patent Laid- 1998) -98047).

However, when nitrogen is added to the crystal from the melt, the nitrogen concentration changes in the longitudinal direction of the crystal due to the segregation phenomenon, which results in a problem of uneven distribution of defects due to the influence.

Further, when nitrogen is added, the defect size is surely reduced and the effect of high-temperature annealing can be facilitated. On the other hand, since the defect density is increased, there is a possibility that deterioration of wafer quality may be adversely affected under insufficient high temperature annealing conditions There is also a problem.

In short, whether or not a wafer obtained by reducing the defect size by the addition of nitrogen (doping nitrogen) is suitable for commercialization is not sufficiently verified, and a wafer doped with nitrogen is used as a wafer for semiconductor device production It is not really known whether it is possible or not.

The present invention relates to a method of manufacturing a silicon wafer for semiconductor device fabrication, which is suitable for non-oxidative heat treatment such as hydrogen-atmosphere high temperature heat treatment (hydrogen heat treatment).

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a graph showing a relationship between an added nitrogen concentration and an oxide film yield yield after hydrogen annealing; FIG.

Fig. 2 is a graph showing the relationship between the added nitrogen concentration and the yield yield of the oxide film when the surface layer of 3 탆 is polished and removed after hydrogen annealing.

3 is a graph showing the measurement results of the constant current TDDB test of the silicon wafer after the hydrogen heat treatment. 3 (A) is a graph showing a measurement result of a constant current TDDB test of a normal element not doped with nitrogen, and FIG. 3 (B) Figure shows the measurement results of the constant current TDDB test of the device.

4 is a graph showing the results of measurement of a gate oxide breakdown voltage (GOI) by a TZDB test of an argon annealed nitrogen doped wafer. 4 (A) is a diagram showing a case in which no polishing is carried out, a case in which polishing after 3 占 퐉 polishing is carried out, and a case in which polishing after 6 占 퐉 polishing are respectively superimposed, and Fig. 4 (B) In order to make it easier to understand the difference from Fig. 2 showing the case after the 탆 polishing, in particular, only the data after 3 탆 polishing is written.

Figures 5A-5D show the results of a constant current TDDB test of an argon annealed nitrogen doped wafer. In particular, Figure 5a to Figure ^{5d, 5 × 10 13 atoms / cm} 3 ( Fig. ^{5b), 1 × 10 14 atoms} / cm 3 ( Fig. 5c), 1.13 × 1O ^14, from any state (Fig. 5a is not doped with nitrogen atoms / cm < ³ > (Fig. 5D) and the nitrogen concentration is increased.

6A to 6D are diagrams showing results of a constant current TDDB test of an argon annealed nitrogen doped wafer. In particular, Figure 6a to Figure ^{6c, 1.64 × 10 14 atoms / cm} 3 ( Fig. ^{6a), 5.82 × 10 14 atoms} / cm 3 ( Fig. ^{6b), 1.37 × 1O 15 atoms} / cm 3 ( Fig. 6c) and the nitrogen concentration And the results are shown in Fig. FIG. 6D is a diagram showing a wafer in which nitrogen is doped at a concentration of 5.82 × 10 ¹⁴ atcms / cm ³ and 1.37 × 10 ¹⁵ atoms / cm ³ , respectively, by hydrogen annealing and argon annealing, respectively.

Fig. 7 is a diagram in which the items in Figs. 5A to 6C are superimposed. Fig.

8 is a graph showing the change in nitrogen concentration in the pulling single crystal when pulled by the CZ method.

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and an object of the present invention is to provide a method of manufacturing a silicon wafer by doping nitrogen with verification of manufacturing conditions capable of producing a silicon wafer having sufficient characteristics for a semiconductor device Doped wafers having desired characteristics as wafers for semiconductor device fabrication after heat treatment is performed.

In view of the above problems, the inventors of the present invention have studied the growth conditions in detail. As a result, they have discovered the conditions for producing silicon wafers having characteristics that are sufficiently common for advanced semiconductor devices even after heat treatment, And has reached the completion of the invention.

More specifically, the present inventors evaluated the wafer characteristics by measuring the gate oxide film breakdown voltage characteristic of a wafer (nitrogen-doped wafer) doped with nitrogen. In this process, the term "nitrogen-doped wafer" The result of the TZDB (Time Zero Dielectrc Breakdown) test tends to be better than that of the wafer not subjected to nitrogen doping. However, when nitrogen is doped at a high concentration, the TDD (Time Dependent Diclectric Breakdown) The results of the tests revealed that anomalies would appear.

The inventors of the present invention also determined whether the result of the TZDB test was good or whether the abnormality resulted from the TDDB test was dependent on the nitrogen concentration of the nitrogen doping wafer but the result of the TZDB test also means that non- The nitrogen concentration at which an abnormality is found as a result of the TDDB test, on the other hand, depends on the kind of the gas (for example, hydrogen gas or argon gas) to be carried out, I got the knowledge that I can guess.

These knowledge contributes greatly to the completion of the present invention. The inventors of the present invention have found that when the heat treatment is carried out in a non-oxidizing atmosphere and the concentration is 4 x 10 ¹⁴ atoms / cm ³ or less Is preferable as a silicon wafer for non-oxidative heat treatment for semiconductor device fabrication for the first time in the present application, and basically this is the content of the right claim.

When the nitrogen doped wafer is evaluated by the gate oxide withstand voltage yield (GOI), the effect of the hydrogen heat treatment is sufficiently exhibited on the surface, and the effect is not related to the nitrogen doping amount. However, In the case where the evaluation is made for the portion, the oxide film pressure-resistance yield rate depends on the nitrogen concentration, and the upper and lower limits of the nitrogen doping amount are present. When the product is used as a semiconductor device product, . In completing the present invention, this knowledge is greatly contributed.

In the above-described Japanese Patent Application Laid-Open No. 10 (1998) -98047, "the nitrogen concentration is at least 1 × 10 ¹⁴ atoms / cm ³ ", but there is no disclosure of information on the basis thereof.

More specifically, the present invention provides the following method and a silicon wafer for non-oxidative heat treatment for semiconductor device fabrication.

(1) the nitrogen concentration is in the range of 5 x 10 ¹³ atoms / cm ³ to 1 x 10 ¹⁵ atoms / cm ³ , preferably in the range of 5 x 10 ¹³ atoms / cm ³ to 8 x 10 ¹⁴ atoms / cm ³ , More preferably in the range of 5 x 10 ¹³ atoms / cm ³ to 4 x 10 ¹⁴ atoms / cm ³ , and still more preferably in the range of 1 x 10 ¹⁴ atoms / cm ³ to 4 x 10 ¹⁴ atoms / cm ³ Silicon wafers for non - oxidative heat treatment for manufacturing semiconductor devices.

That is, when only the gate oxide film pressure by TZDB is taken into account, the range of good products is within the range of 5 × 10 ¹³ atoms / cm ³ to 1 × 10 ⁵ atoms / cm ³ (gate oxide withstand yield of 90 by TZDB (Preferably within a range of 1% to 10%), preferably, the nitrogen concentration is within the range of 1 x 10 ¹⁴ atoms / cm ³ to 8 x 10 ¹⁴ atoms / cm ³ (within the range of the gate oxide film breakdown voltage yield ratio 95% In addition, in consideration of the constant current TDDB, it is required to be 4 × 10 ¹⁴ atoms / cm ³ or less. Consequently, the range of good products is 5 × 10 ¹³ atoms / cm ³ to 1 × 10 ¹⁵ atoms / cm ³ , 5 × 10 ¹³ atoms / cm ³ to 8 × 10 ¹⁴ atoms / cm ³ , Such as a range of 1 × 10 ¹³ atoms / cm ³ to 4 × 10 ¹⁴ atoms / cm ³ , and a range of 1 × 10 ¹⁴ atoms / cm ³ to 4 × 10 20 atoms / cm ³ , Depending on the product to be used.

Examples of the types of the silicon wafers for heat treatment include silicon wafers for heat treatment of hydrogen provided for heat treatment under a hydrogen atmosphere, silicon wafers for heat treatment of argon provided for heat treatment in an argon atmosphere, or silicon wafers for heat treatment in a mixed gas of hydrogen or argon A mixed gas heat treatment silicon wafer, and the like. However, all of the non-oxidizing heat treatment (annealing treatment in an oxygen-free state) for reducing crystal defects in the wafer surface layer are included in the scope of the present invention.

(2) A method for producing a silicon ingot by pulling up a silicon single crystal by the Czochralski method, wherein nitrogen is doped so that the nitrogen concentration becomes 5 x 10 ¹³ atoms / cm ³ to 4 x 10 ¹⁴ atoms / cm ³ Wherein the silicon ingot is produced by pulling up the silicon single crystal under the condition of forming the silicon ingot. Particularly, in a method of manufacturing a silicon ingot by pulling up a silicon single crystal by the Czochralski method, the silicon ingot is doped with nitrogen, and a portion where the nitrogen concentration is 1 x 10 ¹⁴ atoms / cm ³ to 4 x 10 ¹⁴ atms / cm ³ The silicon ingot for producing a silicon wafer for non-oxidative heat treatment is produced.

In one embodiment of the present invention, the silicon wafer for heat treatment for manufacturing a semiconductor device is manufactured by the Czochralski method (CZ method). In this case, nitrogen is doped by a Czochralski method so that the nitrogen concentration of a part or the whole thereof is 5 × 10 ¹³ to 1 × 10 ¹⁵ atoms / cm ³ to pull up the silicon single crystal, A portion having a nitrogen concentration in a range of 5 × 10 ^{13 to} 1 × 10 ¹⁵ atoms / cm ³ , preferably 1 × 10 ¹⁴ to 8 × 10 ¹⁴ atoms / cm ³ is cut out from the silicon ingot, A non-oxidative heat treatment silicon wafer for manufacturing, particularly, a silicon wafer for hydrogen heat treatment or a silicon wafer for argon annealing. When the CZ method is used, a method of applying a magnetic field to the melt (MCZ method) can also be employed.

The method for doping nitrogen includes a method in which nitrogen is mixed into argon gas passing through a furnace during crystal growth, a method in which silicon nitride is dissolved in a raw material melt and pulled up to introduce nitrogen atoms into the single crystal , All currently known methods, and all methods to be found in the future.

Examples of the present invention include the following. A silicon wafer for fabricating a semiconductor device, which is produced by subjecting a silicon wafer for non-oxidative heat treatment for manufacturing a semiconductor device according to the present invention (silicon wafer for non-oxidative heat treatment of the above-described substrate) to hydrogen heat treatment or argon annealing.

A silicon wafer for semiconductor device fabrication in which the doping amount of nitrogen is adjusted in consideration of a virtual element lifetime.

A method of evaluating a wafer doped with nitrogen characterized by determining whether or not the nitrogen-doped wafer can be used as a wafer for semiconductor device fabrication by measuring a virtual device lifetime on the nitrogen-doped heat-treated wafer.

Wherein the method for measuring the virtual element lifetime of the wafer is the TDDB test method.

The "virtual device" means a pseudo-element structure fabricated when the TDDB test method or the TZDB test method is performed, and the "virtual device life time" means that the pseudo device Means the lifetime of the structure.

As experimental examples, the present inventors conducted a hydrogen annealing after cutting a silicon wafer from a CZ-silicon single crystal grown under various conditions, mirror-polishing it, and investigating the defect disappearance effect in the depth direction of the wafer.

<Growth of single crystals>

The decision by adding boron as a dopant 200mm diameter, p-type, crystal orientation <100>, and the V / V × G1 and G2 of the factor (因子) influencing the defective behavior as 0.13~0.4mm ² / min ℃, 0.5 10 &lt; 0 &gt; C / min

(V: pulling rate, G1: temperature gradient in the vicinity of the solid (solid) interface, and G2: temperature gradient in the defect formation temperature range).

&Lt; Doping of nitrogen &

At the time of growing the silicon single crystal ingot, nitrogen was added so that the nitrogen concentration was 4.9 × 10 ^{13 to} 1.24 × 10 ¹⁵ atoms / cm ³ . For comparison, a crystal without nitrogen addition was also prepared.

Here, the addition of nitrogen is carried out by a method of introducing nitrogen gas into the CZ furnace or a method of incorporating a wafer in which a silicon nitride film is formed on the surface into raw silicon (Japanese Patent Laid-Open Publication No. 1993-294780) .

&Lt; Quantitation of nitrogen concentration >

The amount of nitrogen concentration can be quantitatively determined by performing measurement in SINS with respect to a portion where measurement by SIMS is possible and by calculating by calculation using a predetermined calculation expression from a portion measured by SIMS . &Lt; / RTI &gt;

More specifically, by using a SIMS (IMF-6F type, made by Cameca), a crystal lower edge portion having a detection lower limit (a concentration sufficiently larger than ~ 1 x 10 ¹⁴ atoms / cm ³ of a fixed amount of nitrogen %).

In this case, it was confirmed that there was no disturbance in the measurement value due to the residual nitrogen gas from the atmosphere when the sample was exchanged, with the silicon crystal grown by the FZ method not containing nitrogen. Further, the sample was formed by ion implantation of nitrogen. The nitrogen concentration was calibrated with a known standard sample and quantified.

The nitrogen concentration at the position where the measurement by the SIMS can not be performed was calculated by the following calculation formula.

Cs = kCo (1-L) ^k-1

C is the impurity concentration in the crystal, Co is the impurity concentration in the initial melt, L is the degree of solidification, and K is the segregation coefficient of nitrogen. Yatsurugi et al. (1973) J. It was O.0007 from the literature of Electrochem.Soc. 120.975.

The above equation expresses the concentration at which the nitrogen impurity in the silicon melt is taken in by using the segregation coefficient k inherent to the impurity.

In addition, the above equation is an expression showing a segregation phenomenon. According to the segregation phenomenon, the impurity concentration (nitrogen concentration) increases at the same time as the crystal grows, and the concentration becomes high in the crystal tail portion of the crystal, and the detection lower limit (~ 1 x 10 ¹⁴ atoms / cm &lt; ^{3 &} gt ;.

In this respect, in this embodiment, the nitrogen concentration is measured by the SIMS at the tail portion of the crystal straight portion (about 95% of the solidification rate), and the concentration at each crystal position is calculated from the above formula .

<Heat Treatment of Wafer>

The wafer was doped with nitrogen as described above, and subjected to hydrogen heat treatment (hydrogen annealing) and argon annealing. Hydrogen annealing conditions and argon annealing were carried out under the conditions of a general treatment of 1200 DEG C for 1 hour and a state immediately after the hydrogen heat treatment and a state just after the argon annealing and a state after removing 3 mu m of the surface layer after the hydrogen treatment and a state of removing 3 mu m of the surface layer after the argon annealing The gate oxide breakdown voltage characteristic (GCI) was investigated.

<Evaluation of hydrogen annealing (hydrogen annealing) wafer>

[Gate oxide pressure (GOI)]

The gate oxide film breakdown voltage is measured by forming a MOS structure on a wafer and applying a voltage thereto. In the case of measuring the gate oxide film breakdown after polishing by 3 탆, the surface of the wafer is polished by about 3 탆, and a MOS structure is formed thereon, and a voltage is applied to measure it.

[Time Zero Dielectric Breakdown (TZDB) Test]

First, in the measurement of the gate oxide film dielectric breakdown by TZDB, in the measurement method employed in this Experimental Example, as a gate electrode employing the polysilicon electrode of 1Omm ^2, followed by applying a voltage to a step voltage application method. The thickness of the oxide film is 25 nm. The measurement temperature is room temperature (25 占 폚), and the breakdown voltage determination current is 10 占..

As a result, as shown in Fig. 1, the yield of the oxide film pressure-resistant product immediately after the hydrotreating was approximately 100% regardless of the nitrogen concentration and growth conditions.

However, as shown in FIG. 2, when the surface layer 3 m was removed after the hydrotreatment, the surface area of the oxide layer was increased to 6 × 10 ¹⁴ atoms / cm ³ , , And when it is more than that, the yield rate is rather lowered.

Here, in consideration of the completeness of the wafer surface layer (device activation layer), which is increasing in importance as the device is highly integrated, the yield of the oxide film pressure resistance at the depth of 3 탆 from the surface layer should be maintained at least 90%.

2, it is found that the nitrogen concentration needs to be in the range of 5 × 10 ^{13 to} 1 × 10 ¹⁵ atoms / cm ³ in order to make the yield of the oxide film withstand voltage higher than 90% It is found that the nitrogen concentration must be set in the range of 1 x 10 ^{14 to} 8 x 10 ¹⁴ atoms / cm ³ in order to attain the oxide yield voltage of 95% or more.

[Constant current TDDB (Time Dependent Dielectric Breakdown) test]

The constant current TDDB refers to a change in the voltage applied to the element by the breakdown of the oxide film after a predetermined time elapses so that the current becomes constant. In the measurement, a MOS structure is formed as in the inspection of the oxide film breakdown voltage.

In the measurement, in the measurement method employed in this experimental example, a 1 mm ² polysilicon electrode was used as the electrode. The thickness of the oxide film is 25 nm. The measurement temperature is 125 ℃, the applied current density was 50mA / cm ^2, the electric field is determined 4MV / cm.

In the case of measuring the constant current TDDB, a normal element is momentarily destroyed in the vicinity of 100 seconds or more. However, in the case of an abnormal element, instantaneous destruction does not occur, and destruction occurs with time, so that destruction progresses gradually before 100 seconds elapse. Due to this, as a result, it is observed as if breakage has occurred even after 100 seconds have elapsed due to the setting value of the judgment electric field. As a result, when a graph of the measurement result of the constant current TDDB is plotted, a plot is dense in a portion of a normal element when a predetermined time (in this case, about 100 seconds) elapses and a plot is not formed thereon (See Fig. 3 (A)), and in the case of an abnormal element, a plot is also formed at a position apart from the elapsed time of 100 seconds other than the above (see Fig.

The occurrence of such an anomaly in the heat-treated wafer is not observed when the wafer not doped with nitrogen is heat-treated (see Fig. 3 (A)), and a phenomenon unique to the heat treatment of a wafer doped with nitrogen at a high concentration to be.

In addition, the TDDB test is a kind of reliability test of a semiconductor device, and is an index of the steady-state resistance of the device to fatigue. The TDDB test may be regarded as providing virtual data for estimating the stability of the life of the semiconductor device manufactured from the wafer to be tested.

In this experimental example, the constant current TDDB in the un-polished state and the constant current TDDB in the polished 3 μm were measured for each of the wafers after the hydrogen annealing.

As a result, the data shown in the following table was obtained.

[Table 1]

[Table 2]

Table 1 shows the relationship between the nitrogen concentration and the constant current TDDB abnormality rate in the non-polishing state, and Table 2 shows the relationship between the nitrogen concentration and the constant current TDDB abnormality rate after 3 占 퐉 polishing. As shown in these tables, the constant current TDDB shows a normal value when the nitrogen concentration is 4 x 10 ¹⁴ atoms / cm ³ or less. Therefore, by adding this requirement to the conditions of the oxide film pressure, the nitrogen-doped crystal optimum condition means that the nitrogen concentration is within a range of 5 x 10 ¹³ atoms / cm ³ to 4 x 10 ¹⁴ atoms / cm ³ As the silicon wafer for heat treatment, particularly, a silicon wafer for heat treatment having a nitrogen concentration falling within the range of 1 x 10 ¹⁴ atoms / cm ³ to 4 x 10 ¹⁴ atoms / cm ³ .

&Lt; Evaluation of argon annealing wafer &

[Time Zero Dielectric Breakdown (TZDB) Test]

The gate oxide breakdown voltage (GOI) by the TZDB test was measured under the same conditions as in the case of the hydrogen annealing. In the case of argon annealing, however, the gate oxide film breakdown after polishing with 6 mu m was also measured. The results are shown in Fig. 4 (A), 4 (A), 5 (B) and 5 (B) show the case where no polishing, the case after 3 탆 polishing and the case after 6 탆 polishing are superimposed respectively, In order to make it easier to understand the difference from Fig. 2 showing the case of 3 占 퐉 polishing, only data after polishing of 3 占 퐉 are written.

As apparent from Fig. 4, the nitrogen doped wafer exhibits excellent gate oxide breakdown voltage characteristics in accordance with the TZDB test and has a definite upper limit such as the wafer after 3 탆 polish after hydrogen annealing Is not observed. Thus, the argon anneal to the wafer doped with nitrogen, TZDB is not present to the same upper limit to the nitrogen concentration On examination, or significantly more than the upper limit of the nitrogen concentration 1 × 10 ¹⁵ atoms / cm ³ in the case of hydrogen annealing top of There is an upper limit value. As is apparent from Fig. 4 (A), argon annealing is not much different from the result of the TZDB test, either after 3 μm polishing or after 6 μm polishing.

[Constant current TDDB (Time Dependent Dielectric Breakdown) test]

The constant current TDDB test was also performed under the same conditions as in the hydrogen annealing as in the TZDB test. The results are shown in Figs. 5A to 7.

5A to 5D and FIGS. 6A to 6D show a state in which nitrogen is not doped at all (5 × 10 ¹³ atoms / cm ³ (FIG. 5B), 1 × 10 ¹⁴ atoms / cm ^{3 , 1.13 × 10 14 atoms / cm} 3 ( Fig. ^{5d), 1.64 × 10 14 atoms} / cm 3 ( Fig. ^{6a), 5.82 × 10 14 atoms} / cm 3 ( Fig. ^{6 b), 1.37 × 1O 15} atoms / cm 3 ( 6c) and the results of the constant current TDDB test when the nitrogen concentration is increased.

As apparent from these figures, no change occurs in the dense state of the plot until the nitrogen concentration is 1.64 x 10 ¹⁴ atoms / cm ³ (FIG. 6 a), but the nitrogen concentration is 5.82 × 10 ¹⁴ atoms / cm ³ At 1.37 x 10 &lt; ¹⁵ &gt; atoms / cm &lt; ³ &gt; (Fig. 6C), deviations of the plots peculiar to the ideal device are seen in much later than 100 seconds.

FIG. 6D shows a wafer in which nitrogen is doped at a concentration of 5.82 × 10 ¹⁴ atoms / cm ³ and 1.37 × 10 ¹⁵ atoms / cm ³ , and the wafer is subjected to hydrogen annealing and argon annealing, respectively, in a superimposed manner. 7A and FIG. 7B show the overlapping display of FIGS. 5A to 6C.

From these results, it can be seen that nitrogen doped wafers were subjected to argon annealing at a nitrogen concentration of 1.64 × 10 ¹⁴ atoms / cm ³ (FIG. 6A) to 5.82 × 10 ¹⁴ atoms / cm ³ , Which is close to the value of 4 x 10 ¹⁴ atoms / cm ³ (Fig. ⁶ (d)) at the upper limit on the result of the constant current TDDB test for hydrogen annealed.

In this respect, it can be seen that there is a large difference in the results of the TZDB test with respect to hydrogen annealed and argon annealed nitrogen doped wafers, but there is almost no difference in the results of the constant current TDDB test.

As the silicon wafer for semiconductor devices (silicon wafer for semiconductor device fabrication), the results obtained by the TDDB test must also be taken into consideration. Therefore, as nitrogen-doped wafers for non-oxidative heat treatment to be subjected to hydrogen annealing or argon annealing, As a result of the test, it can be seen that nitrogen should be contained at the upper limit of 4 x 10 ¹⁴ atoms / cm ³ or less.

8 is a graph showing changes in the nitrogen concentration in the pulling single crystal when pulled up by the CZ method. 8, the abscissa indicates the solidification ratio when the entire amount of the polysilicon material, which is the material of the single crystal ingot, is 1, and the ordinate indicates the nitrogen concentration. In the figure, (a) when a curve in the case where the initial concentration of 2 × 10 ^14, (b) is a curve in the case where the initial concentration 1 × 10 ^14, (c) is the initial concentration of 5 × 10 ¹³ Respectively.

As apparent from Fig. 8, the tail side of the pulled-up silicon ingot (the shape of the corresponding silicon ingot corresponding to the graph is shown) becomes higher in nitrogen concentration due to the segregation phenomenon of nitrogen.

INDUSTRIAL APPLICABILITY As described above, according to the present invention, it is possible to provide a silicon wafer doped with nitrogen that exhibits excellent wafer characteristics as a product.

Claims (8)

Wherein the nitrogen concentration is in the range of 5 x 10 ¹³ atoms / cm ³ to 1 x 10 ¹⁵ atoms / cm ³ . Wherein the nitrogen concentration is in the range of 5 x 10 ¹³ atoms / cm ³ to 4 x 10 ¹⁴ atoms / cm ³ . 3. The method according to claim 1 or 2, Silicon wafer for non-oxidative heat treatment for semiconductor device fabrication which is a silicon wafer for hydrogen heat treatment or a silicon wafer for argon annealing. A method of manufacturing a silicon ingot by pulling up a silicon single crystal by a Czochralski method, Nitrogen was doped to form a silicon ingot for forming a non-oxidative heat treatment silicon wafer, under the condition of forming a portion having a nitrogen concentration of 5 × 10 ¹³ atoms / cm ³ to 1 × 10 ¹⁵ atoms / cm ³ How to. A silicon wafer for manufacturing a semiconductor device, which is produced by subjecting the silicon wafer for non-oxidative heat treatment according to any one of claims 1 to 3 to hydrogen heat treatment or argon annealing. A silicon wafer for semiconductor device fabrication in which the doping amount of nitrogen is adjusted in consideration of a virtual element lifetime. A method of evaluating a wafer doped with nitrogen characterized by determining whether or not the nitrogen-doped wafer can be used as a wafer for semiconductor device fabrication by measuring a virtual device lifetime on the nitrogen-doped heat-treated wafer. 8. The method of claim 7, Wherein the method for measuring the virtual lifetime of the wafer is a TDDB test method.